High Efficient Dye-Sensitized Solar Cells Based on Synthesized SnO 2 Nanoparticles

In this study, SnO2 semiconductor nanoparticles were synthesized for DSC applications via acid route using tin(ii) chloride as a starting material and hydrothermal method through the use of tin(iv) chloride. Powder X-ray diffraction studies confirmed the formation of the rutile phase of SnO2 with nanoranged particle sizes. A quasi-solid-state electrolyte was employed instead of a conventional liquid electrolyte in order to overcome the practical limitations such as electrolyte leakage, solvent evaporation, and sealing imperfections associated with liquid electrolytes. The gel electrolytes were prepared incorporating lithium iodide (LiI) and tetrapropylammonium iodide (Pr4N I) salts, separately, into themixture which contains polyacrylonitrile as a polymer, propylene carbonate and ethylene carbonate as plasticizers, iodide/triiodide as the redox couple, acetonitrile as the solvent, and 4-tertiary butylpyridine as an electrolyte additive. In order to overcome the recombination problem associated with the SnO2 due to its higher electronmobility, ultrathin layer of CaCO3 coating was used to cover the surface recombination sites of SnO2 nanoparticles. Maximum energy conversion efficiency of 5.04% is obtained for the device containing gel electrolyte incorporating LiI as the salt. For the same gel electrolyte, the ionic conductivity and the diffusion coefficient of the triiodide ions are 4.70 × 10 S cm and 4.31 × 10 cm s, respectively.


Introduction
Dye-sensitized solar cells (DSCs) have been investigated as the next generation of solar cells, due to their low production cost and easy fabrication procedures compared to those of conventional silicon solar cells.In 1991, Grätzel and his coworker invented a solar cell based on the ruthenium sensitizer adsorbed on nanoporous TiO 2 semiconductor film [1].However, the charge separation ability of TiO 2 -based DSC is suppressed by its low electron mobility (<1 cm 2 V −1 s −1 ) resulting in a higher dark current [2].Moreover, TiO 2 shows a high photocatalytic effect and, as a result, sensitizer which is attached onto the nanoporous TiO 2 network, prone to degrade rapidly.Therefore, in order to overcome these problems regarding TiO 2 -based DSCs, other high band gap semiconductors such as ZnO, SnO 2 , and CdS were investigated as possible alternative semiconductor materials.Among these semiconductor materials, SnO 2 shows relatively high electron mobility (∼250 cm 2 V −1 s −1 ), high electron-hole separation ability, and high transport properties [3,4].Other than its attractive properties, SnO 2 nanoparticles can be synthesized using various techniques such as the hydrothermal method, sol-gel, and acid route.Here, we mainly aim to fabricate DSCs-based on the as-prepared SnO 2 nanoparticles using acid route method and the hydrothermal method.Although SnO 2 has promising electronic properties such as higher electron mobility, it shows inferior performance due to recombination of injected electrons with excited dye molecules and redox species of electrolyte.The recombination flux can be naturally divided into three elements as follows: (1) Electron reaching recombination sites in the semiconductor surface.
(2) Ions or hole reaching the surface from the electrolyte or hole-conductor side.
Somehow, recombination occurs from the conduction band energy levels,  cb  , and surface trap levels,  ss  , can be expressed as follows: where the terms have their usual meanings.Generally, recombination rate is mainly affected by two factors: the position of the semiconductor energy level with respect to the redox levels and treatment of surface which intercept charge transfer from the semiconductor.
In this study, we focus on the blocking the semiconductor surface trap levels through the use of very thin layer of high band gap insulating material.Since the efficient electron tunnelling should be on the order of less than 10 nm, study is focused on employing an ultrathin layer of coating layer on SnO 2 -based DSCs without greatly reducing the rate of photoinjection from excited dye molecules, that is, forward tunnelling of electrons.In order to verify the probability of forward tunnelling and back tunnelling of injected electrons, tunnelling transmission coefficient, , equation was employed.
where  0 is the barrier height,  is the effective electron mass, and  is the barrier width.Photoexcited dye molecules cannot inject the electrons to the higher energy level of conduction band of the insulating material (band gap 6.00 eV-9.00 eV) which is greater than 2.5 eV with respect to the Lowest Unoccupied Molecular Orbital (LUMO) level of the dye molecules.The only possible way of reaching excited electron to conduction band of the semiconductor is by tunnelling.After the electrons tunnel into the conduction band, they relax towards the bottom of the conduction band acquiring lower energy state.These electrons will see the energy difference between the conduction band of the semiconductor and conduction band of the insulating material as a barrier height,  0 (>4.3 eV).Then back tunnelling of electrons will be greatly reduced compared to forward tunnelling thus keeping the Fermi level of SnO 2 in equilibrium.
Therefore, in order to suppress the recombination which occurs at the electrolyte/semiconductor interfaces of the SnO 2 semiconductor network, CaCO 3 coating layer is introduced to the system [5].This CaCO 3 layer would increase the photovoltaic performances of the composite SnO 2 /CaCO 3based DSCs as it acts as a barrier to the electrons which are in the conduction band of the SnO 2 semiconductor network, by suppressing the back tunnelling of electrons.Moreover, the conventional DSCs which consist of liquid electrolytes suffer from practical limitations, such as electrolyte leakage, solvent evaporation, and sealing imperfections.The use of gel electrolytes instead of liquid electrolytes would help to circumvent the above-mentioned drawbacks to a certain extent as gel electrolytes have promising properties, such as thermal stability, nonflammability, and nonvolatility.In this study the quasi-solid-state dye-sensitized solar cells (QSDSCs) were fabricated using as-prepared SnO 2 /CaCO 3 composite working electrode and the gel electrolyte consists of propylene carbonate and ethylene carbonate as plasticizers, polyacrylonitrile as a polymer, acetonitrile as the solvent, 4tertiary butylpyridine as an electrolyte additive, and lithium iodide and tetrapropylammonium iodide as salts [6].Additionally, this study was carried out with a metal-free organic dye, namely, indoline D358 dye, which has a high chelating ability ( = 13000).
In order to study the size effects of the nanoparticles used in DSC fabrication, SnO 2 nanoparticles of different size were synthesized in our laboratory.Particle size could affect the device performance in two different ways.When the particle size is reduced to nanoscale, the effective surface area increases about 1000 times, hence increasing the dye adsorption by the same factor.
The effective surface area of a particle is increased when the particle size is decreased.For a given volume the relationship between the effective surface area () and the particle size () can be expressed as follows [7]: Here  is the density of the material and  is a constant for the material of interest.The effective surface area of the material increases intensely for smaller particles resulting in higher dye adsorption.This is a very positive contribution to enhance device performance.On the other hand, the lower particle size also leads to increased leaking of electrons from the semiconductor to recombine either with the redox species or with the oxidized dye molecules.It is explained as follows.Electron transport from the semiconductor network to FTO depends on these trapping and detrapping processes [8][9][10].The following wave function describes the electrons in shallow traps [5]: where  is the radial coordinate measured from the trapped site and  is the parameter which has dimensions of length and it can be given as follows: where  * is the effective electron mass.At room temperature, the parameter  is ∼4 nm which is in the same order of magnitude as the crystallite radius of SnO 2 .That means electrons that are in the trapped levels could easily leak into the electrolyte.
The electrons leakage and the amount of dye absorption are competing effects on the performance of DSCs.Therefore, we try to minimize the electron leakage though the use of SnO 2 /CaCO 3 composite system by sacrificing the dye loading to some extent.

Preparation of SnO 2 Nanoparticle
Acid Route.First, tin(ii) chloride (1.71 × 10 −2 mol, Aldrich, 98%) and citric acid (4.16 × 10 −2 mol) were dissolved in deionized water until a saturated aqueous solution was formed.The resultant solution was heated slowly on a hot plate at 80 ∘ C to evaporate the solvent.Here, an amorphous glassy material was formed after the complete removal of water.Next, the amorphous material was calcined in a furnace in air for 30 minutes by varying temperature from 500 ∘ C to 750 ∘ C.
Hydrothermal Method.The initial solution was prepared using the SnCl 4 ⋅2H 2 O (3.4 g, Aldrich, 98%), conc.HCl (8.00 cm 3 , Aldrich, 98%), and ethanol (40.00 cm 3 , Aldrich, 98%).This ratio gives the maximum yield as found by the preliminary study.The solution was poured into a Teflon flask and autoclaved at different temperatures from 180 ∘ C to 250 ∘ C for 15 hours.The above temperature range was selected since it was found that, at temperatures below 150 ∘ C, the required material is not formed and temperatures between 150 ∘ C and 180 ∘ C do not give a considerable yield.Next, the greenish white precipitate was washed several times with deionized water (to remove excess ions in the medium), followed by ethanol.Then the precipitate was dried in a vacuum oven at 60 ∘ C for 24 hours.

Fabrication of Dye-Sensitized Solar Cells.
As-prepared SnO 2 (0.60 g), acetic acid (Aldrich, 98%, 10 drops), Triton X-100 (Aldrich, 98%, 3 drops), ethanol (Aldrich, 98%, 40.0 cm 3 ), and CaCO 3 (0.040 g, Aldrich, 98%) were mixed thoroughly and the resulting SnO 2 /CaCO 3 suspension was used to make devices after undergoing ultrasonic treatment.The SnO 2 /CaCO 3 suspension was sprayed onto well-cleaned FTO glass (10 Ω cm −2 ) plates heated to 150 ∘ C on a hot plate.Then, the samples were sintered at 500 ∘ C for 30 minutes and were allowed to cool down to 80 ∘ C. The samples were then immersed in an indoline D358 dye solution for 12 hours, and the dye coated-SnO 2 /CaCO 3 films were rinsed with acetonitrile to remove any physically adsorbed dye molecules.Next, the electrolyte was sandwiched between the FTO/SnO 2 /CaCO 3 working electrode and a lightly platinized FTO counter electrode (∼7 Ω/sq, Aldrich) to assemble the solar cell device.Same procedure was followed to prepare the SnO 2 -based DSCs.

Preparation of Gel Polymer Electrolyte.
In this experiment, 0.225 g of polyacrylonitrile (Aldrich), 0.525 g of ethylene carbonate (Aldrich, 98%), 0.750 g of propylene carbonate (Aldrich, 99%), and 0.020 g of iodine (Aldrich, 98%) were mixed and stirred well in a magnetic stirrer for 12 hours.0.150 g of LiI (Aldrich, 99%) (electrolyte Y) and 0.150 g of Pr 4 N + I − (Aldrich, 98%) (electrolyte X) were used separately to prepare the gel electrolyte.Each time, the electrolytes were stirred at 80 ∘ C until the mixture turned into a clear, homogeneous, viscous gel.In each case, the gel electrolytes were subsequently pressed by sandwiching them between two clean glass plates to obtain a free-standing polymer film.They were subsequently dried in a vacuum desiccator overnight, at room temperature, to remove any absorbed moisture.

Characterization.
In order to study the film morphology and the performance of the solar cell device, the following characteristic techniques were conducted.Crystallographic characterization of the SnO 2 powder and composite SnO 2based films were done by means of powder X-ray diffraction (XRD) using a Siemens D5000 X-ray diffractometer with the Cu K radiation ( = 0.1540562 nm) at a scan rate of 1 ∘ min −1 .The scanning electron microscopic (SEM) images were obtained using a Joel 6320 F scanning electron microscope.UV-Visible spectroscopy studies were carried out using a UV-spectrophotometer (UV-1800, SHIMADZU).The photovoltaic performance of the DSCs was measured by a solar simulator (PECCELL PEC-L01) with a source meter (Keithley 2400) at 25 ∘ C under AM 1.5 (100 mW cm −2 ) illumination.The total irradiated area of the DSCs was 0.25 cm 2 .In order to determine the particle size of the newly prepared SnO 2 powder, a particle size analyzing experiment was carried out using a particle size analyzer Cilas Nano DS.The Electrochemical Impedance Spectroscopy (EIS) studies were carried out with a potentiostat (PGSTAT12) with a forward bias of −0.58 V under dark conditions.The AC signal was ±10 mV in the frequency range of 0.01 Hz to 10 kHz.

Results and Discussion
First, newly prepared SnO 2 , using the acid route, was characterized and then moved onto the hydrothermal route which is rather of low cost.Initially, we attempted to find the better mediator for the synthesis of SnO 2 .Here, malonic acid, malic acid, and citric acid were employed to prepare SnO 2 powder at 500 ∘ C and 600 ∘ C. Then the resultant SnO 2 powder was used to fabricate the DSCs.Results obtained are tabulated in Tables 1 and 2.
Results in Tables 1 and 2 showed that the nanoparticles prepared with citric acid gave the best performance.This is possibly due to citric acid acting as the fuel itself giving out a large amount of heat during combustion.Combustion of the metal-acid complex is highly exothermic and releases a large amount of heat for quick conversion of the complex into its metal oxides.When citric acid is employed, there is no need to supply extra fuel into the system.The other possible reason is the formation of bigger particles when we employ malic acid and malonic acid in place of citric acid.The possible complex of a divalent metal cation (Sn 2+ ) and citric acid can be depicted as shown in Figure 1 [11,12].
If we consider the steric properties of malic and malonic acid, there is a higher possibility of aggregating more Sn 2+ cations when malic and malonic acids are used as a mediator compared to the bulky structure of citric acid as shown in Figure 1.This hypothesis supports the argument that preparation of SnO 2 nanoparticles using citric acid route is more efficient.
In order to find the average particle size of SnO 2 , initially we conducted the XRD studies and further confirmation was carried out using the particle size analyzer.
Next, the particles were dispersed in deionized water by a magnetic stirrer and sonicated several times until a transparent colloidal solution is formed.Then the measurements were carried out using a particle size analyzer and the results obtained are shown in Figure 3.
According to the above plot, the sizes are much larger than expected as to the total diameter measured with the hydronium ions which were surrounded by SnO 2 nanoparticles.As the isoelectric point is very low (∼5 pH) in SnO 2 , it is easy to attach many hydronium layers on the SnO 2 nanoparticles.It might be the reason for the higher hydrodynamic diameter compared to the particle size calculated by the XRD data as shown in Table 3.According to the results given in Table 3, particle sizes increase with the temperature.This might be due to the fact that the higher temperature will help to sinter with neighbour SnO 2 nanoparticles thus making bigger particles.
Table 4 gives the variation of solar cell parameters with the particle size.Out of the six temperature values, SnO 2 nanoparticles synthesized at 650 ∘ C gave the best performance.According to the results shown in Table 3, SnO 2 prepared in 650 ∘ C gave 37 nm which is the middle of the particle sizes in between 25 nm and 53 nm.DSCs prepared in 550 ∘ C to 650 ∘ C showed similar  SC values.This might be due to the fact that the particle size of about 30 nm is sensitive to both negative and positive effects occurring due to recombination and dye attachment. OC obtained for the DSCs fabricated at higher temperature is quite law.Even though we expected higher  OC values from those DSCs due to reduction of the recombination as the bigger particle reduces recombination, it seems that bigger particle size greatly reduces the absorbed dye amount compared to the reduction of recombination.
In order to analyze the reasons for the variation of solar cell parameters with the particle size, our next attempt was to investigate the variation of the amount of dye loading with the particle size distribution as shown in Table 3.As previously discussed, samples were prepared in 1 cm 2 area for desorption of the dye and the results obtained are plotted in Figure 4.
It is obvious that the highest dye amount can be observed for the particles with the lowest size.But the lowest particle size did not give the maximum efficiency.If it shows the highest dye loading, then it would be rich in electrons which are received from photoexcited dye molecules.The same fact will lead to increase in recombination as higher electron density increases the driving force of electrons.
But, again, lower particle size means that it is easier for the electrons to come out to the surface of SnO 2 nanoparticles and recombine with triiodide or excited dye molecules as discussed in the Introduction.That would be the possible reason for lower performance of the lowest particles size.In order to confirm these explanations, EIS studies were  conducted and calculated effective diffusion length together with the electron lifetime is given in Table 5.
The equivalent circuit was used to obtain the resistance and capacitance values which are presented in Table 5.Smaller observed   value means that the conduction band electrons could easily recombine with the triiodide ions in the electrolyte, thus lowering  OC .  values obtained for the device fabricated with particles synthesized at 500 ∘ C are much smaller compared to the device fabricated with particles synthesized at 650 ∘ C. The electron recombination occurring in particles sensitized at 500 ∘ C and 750 ∘ C could be much faster compared to the particles sensitized at 650 ∘ C.This can be further confirmed by using the electron lifetimes  which were obtained from EIS by fitting the experimental data through an appropriate equivalent circuit.Device fabricated using SnO 2 nanopowder synthesized at 650 ∘ C showed 125% of increase when compared to device made by employing nanoparticles synthesized at 500 ∘ C. The next attempt was to systemize the SnO 2 nanoparticles using the hydrothermal method.As discussed earlier, this is a very simple and cheap method.SnCl 4 , HCl, and ammonia were used in a precursor solution and it was autoclaved under relatively low temperature.The resultant greenish powder was separated and the precipitate was washed several times with deionized water and vacuum dried for 12 hours.Thus the powder samples obtained underwent XRD characterization and results are tabulated in Table 6.Samples were also used to fabricate DSCs and results are given in Table 7. Table 6 also shows the same trend of Table 3.This might be due to the same reason described in earlier section.
In the hydrothermal method, formation of SnO 2 nanoparticles can be explained as follows.The behavior of NH 3 may correlate to a process named molecule recognition that could have taken place at the inorganic/organic interface due to charge and stereochemistry complementarity [13,14].In an aqueous medium, NH 3 would ionize completely and result in a tetrahedral orientation with the electron lone-pair and it would be incorporated with Sn 4+ cation as depicted on the left side of Figure 5.A possible mechanism for the tinplating process is forming micelles that contained many tin cations (H 3 N-Sn 4+ complexes) on the surface.The micelles act as nucleating points for the growth of SnO 2 crystals.During the hydrothermal process, H 3 N-SnO 2 complexes could be formed.And they coalesce to form a large particle.Since the crystallization process was under the critical control of NH 3 , resulting particles were invariably spherical [15,16].Generally, coagulation of the submicron sized particles only occurred in the sample prepared at higher precursor concentrations as NH 3 molecules were unable to fully envelop the particles while crystal growth took place.Therefore, low precursor concentrations were used here in order to control the aggregation of particles.
Cells prepared with only SnO 2 showed relatively low performance due to lower  OC and fill factor as shown in Table 8.In order to enhance  OC of the solar cell devices, CaCO 3 coated SnO 2 dye-sensitized solar cells were employed.As one of the prime aims of this study is to find the practical suitability of a gel polymer electrolyte in DSCs, here we employed a gel polymer electrolyte.As expected,  OC is enhanced while sacrificing the short-circuit current density to a certain extent as depicted in Figure 6.
In order to examine the composite nature of the fabricated films, XRD studies were conducted.Figure 7 shows XRD spectra obtained for all three composite materials and the composition is verified using standard ICCD data.The planes responsible for these diffractions due to each compound are shown within the XRD spectra.The scanning electron microscopic (SEM) graphs were examined in order to study the film morphology.According to Figure 8, they show interconnected porous structures in composite systems which are favorable for enhanced dye adsorption due to increase in the surface area and for sufficient electrolyte penetration due to nanoporous structures.
Table 9 shows the performance of the solar cell devices fabricated using hydrothermal method.According to the results shown in Tables 8 and 9, CaCO 3 coated SnO 2 device shows 55% increase of efficiency compared to the bare SnO 2based device.It is considerable amount increase.This can be considered to be due to reduction of the recombination as coating layer suppresses the back tunnelling of the injected electron.This phenomenon will be proved by the 30% increase of  OC due to CaCO 3 coating on top of the SnO 2 nanoparticles.As shown in Table 9, the best - performances were obtained for the electrolyte  due to the higher short-circuit current density compared with that of .This might be due to the lowering of conduction band edges by shifting towards a more positive potential via adsorbing Li + ions onto the semiconductor surface.Therefore, a favorable energy gap will be formed for the electron injection from the sensitizer molecules to the conduction band of the SnO 2 and thereby  SC of the device is increased while lowering  OC of the device compared to the electrolyte  based DSCs.
Another possibility for the higher  SC value of the device fabricated with the electrolyte  is the formation of a more amorphous polymer network which helps transport triiodide ions.Coordination interaction in between CN groups of PAN and Li + ions will help formation of cross-linking site thus increasing amorphousness of gel polymer electrolyte [6].

Conclusion
The SnO 2 nanoparticles were synthesized using tin(iv) chloride and tin(ii) chloride as the starting materials.The best particle size for DSCs applications is of about 30 nm.Even though SnO 2 -based DSCs show higher current density they show a low open circuit voltage.One possible method of improving  OC is the introduction of CaCO 3 coating layer sacrificing the current density to a certain extent.Comparably the best photovoltaic performances were obtained with the gel electrolyte consisting of LiI for the composite SnO 2 /CaCO 3 system.The highest values of  SC ,  OC , fill factor, energy conversion efficiency, ionic conductivity, and diffusion coefficient of triiodide ions with LiI-based gel electrolyte, were 17.7 mA cm −2 , 0.573 V, 0.496, 5.04%, 4.70 S cm −1 , and 4.31 × 10 −7 cm 2 s −1 , respectively.These results could be attributed to the higher degree of amorphous nature of the gel electrolyte, due to formation of cross-linking sites with Li + ions and due to the suppression of the recombination by TBP and the CaCO 3 coating layer.

Figure 1 :Figure 2 :
Figure 1: Chemical structures and possible formation mechanism of SnO 2 .

Figure 3 :Figure 4 :
Figure 3: Particle size distribution with the calcination temperature.

Figure 5 :Figure 6 :
Figure 5: Formation of nucleation and crystal growth mechanism.

Table 1 :
parameters of the devices fabricated with as-prepared SnO 2 synthesized at 500 ∘ C.

Table 2 :
parameters of the devices fabricated with as-prepared SnO 2 synthesized at 600 ∘ C.

Table 3 :
Particle sizes calculated using XRD pattern and particle size analyzer.

Table 4 :
parameters of the device fabricated with as-prepared SnO 2 and with the Li-based gel electrolyte.

Table 5 :
EIS parameters of the device fabricated with as-prepared SnO 2 which is synthesized by varying the temperature.

Table 6 :
Variation of particle size with the autoclave temperatures.

Table 7 :
parameters of devices fabricated with as-prepared SnO 2 and with the liquid electrolyte.

Table 8 :
parameters of the devices fabricated using different synthesization techniques.